Connect with us

Technology

YouTube Recommendation Algorithm Remains Threat Against Pedophiles

YouTube recommendation algorithm allows pedophiles to find videos of children in compromising situations Click To Tweet

Published

on

Photo by Ludovic Toinel on Unsplash

Barely three months after YouTube exposed as a “soft-core pedophile ring” in February, The New York Times reports today that the issue remains persistent through its recommendations algorithm.

YouTube is a massive social media platform that offers hours and hours of video content uploaded by millions of people from all over the world through individual accounts called “channels.” It’s an FTC-approved platform where people can watch videos according to their liking. However, it has also become a source of disturbing intentions like pedophiles who flocked the platform with contents that aim fulfill their fetishes.

In February, YouTuber Matt Watson uploaded a video where he detailed how the “soft-core pedophile ring” operated in YouTube. He explained that pedophiles posted initials CP for child pornography and video time stamps to allow others to skip to seemingly innocuous parts of videos of children playing in a pool or doing a “bikini haul.”

“On its own, each video might be perfectly innocent, a home movie, say, made by a child. Any revealing frames are fleeting and appear accidental. But, grouped together, their shared features become unmistakable,” Times states.

The timestamps shared by CP accounts show parts of the videos with exposed genitals can be seen, or when a child is in a compromised position such as doing splits or accidentally lifting her shirt too high and exposing her nipples. Some girls in these videos are as young as five years old but their videos amass views reaching hundreds of thousands, if not millions, with even more comments.

In February, YouTube has immediately acknowledged the severity of the issue and enacted to address the problem by deleting hundreds of thousands of accounts relating to pedophilia and has banned the comments section on videos with children in it.

“YouTube’s terms of service state that children under the age of 13 aren’t allowed to have their own accounts, but many of these innocuous videos are uploaded by older family members. Many children are also key components to an entire genre on YouTube known as “family vlogging.” Creators like The Ace Family (16.4 million subscribers), Tydus and Cor (2.8 million subscribers), Daily Bumps (4.6 million subscribers), and Roman Atwood Vlogs (15.2 million subscribers), put their kids front and center,” The Verge explains.

YouTube’s initial decision to ban the comments section with videos involving minors affected the majority of “family vloggers.” The traffic or number of comments in a video is put into consideration when the platform’s recommendation algorithm decides which videos a viewer might want to watch next.

YouTube’s automated recommendation system — which drives most of the platform’s billions of views by suggesting what users should watch next —is part of a phenomenon on the platform called a wormhole effect, as Watson calls it.

Mainly, YouTube hooks its viewers by recommending related but different videos to the one you’re currently watching. For example, a person could be watching an Ace Family video about how they get their kids to get ready for school — YouTube would then recommend kids playing in a park. In other themes, this would lead viewers to continually consume more content without getting tired of watching a single dimensional set of videos.

The moment YouTube decides to recommend seemingly random videos of children, it becomes a moment for pedophiles to take advantage. According to Time, YouTube has unconsciously created a convenient archive of children in various compromising situations for pedophiles to consume. Parents upload these videos with pure intentions such as sharing how their kids are enjoying the moment — much like a home video.

“It’s YouTube’s algorithm that connects these channels,” said Jonas Kaiser, one of three researchers at Harvard’s Berkman Klein Center for Internet and Society who stumbled onto the videos while looking into YouTube’s impact in Brazil. “That’s the scary thing.”

Specifically, it’s not the pedophiles who point which videos they should watch — its YouTube’s algorithm who’s responsible for suggesting videos that are seemingly popular with other pedophiles, most of which have hundreds of thousands of views and dozens of disturbing comments.

The extraordinary view counts and comments indicated that the system found an audience for the videos, and was “keeping that audience engaged,” says Yasodara Córdova, a researcher who has studied the distribution of online pornography.

On the other hand, YouTube completely disabling the recommendations feature for channels with minors in their videos would affect family vloggers the most. Established channels could potentially lose millions of views. Meanwhile, up and coming family-oriented channels could find it hard to grow on YouTube — being a YouTube creator has developed to become a source of income for some.

YouTube is currently working on a balance that would eliminate threats posed by pedophiles all the while considering the position of family vloggers in their platform.

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Technology

Facebook Live Fail: Pakistan Minister In Cat Filter

Another Facebook Live fail: Pakistan’s Minister of Information and three other officials turned into cartoonish cats when volunteer accidentally turns on Facebook’s Camera Effects feature.

Published

on

Facebook Live
Facebook Live was used to document Pakistani Minister's Press Briefing last June 14, 2019. Photo: Sticker Mule | Unsplash.com

A serious government press meeting has turned into the newest internet meme. Last Friday, Shaukat Ali Yousafzai, 10th Khyber Pakhtunkhwa Assembly’s Minister of Information conducted a press briefing looking like a cat. Accidentally, a cat filter was applied on Press Trust India’s (PTI) Facebook Live session coverage.

Viewers saw various government officials with pink cat ears and cartoonish whiskers. It took a few minutes before the representatives of PTI realized the mistake, despite prompts from viewers via comments on the video. After the press briefing, PTI immediately deleted the video. However, screenshots of the incident have now gone viral.

Mr. Yousafzai addressed the amusing event through a statement to the AFP news agency. He says, “I wasn’t the only one – two officials sitting along me were also hit by the cat filter.”

PTI has since released a statement through Twitter, identifying that the mistake was done by a hardworking volunteer.

In Facebook Live, applying filters is one click away. Clicking on the magic wand at the lower left corner of your phone’s screen will enable different kinds of templated masks, borders, and other effects to your video. The first option on the screen would be the most recent filters you have used.

Facebook Live on Important Events

According to a report, there are nearly two billion users per month who watched Facebook Live sessions in 2017 from its launch in 2016. It has since then become a tool favored by users. Since then, users have used Facebook Live to share important events such as weddings, birthday parties, and family reunions.  

With this kind of user engagement, governments are also noticing Facebook Live’s potential in reaching out to their voters. The Pakistan Government is not the only one using Facebook Live to stream official proceedings. Local and National governments use this tool to engage with their constituents through broadcasting events such as the signing of a bill into law and debate proceedings in the US Senate, to name a few.  

Facebook has since then developed additional tools related to public policies. This is what Katie Harbath, Facebook’s Public Policy Director, presented in the 2016 Government Social Media Conference held in Reno, Nevada. During the event, Harbath also shared that a good number of viewers in a Facebook Live session about Canada’s budgetary hearings; something that wasn’t expected at that time. In the 2016 US elections, the live-streaming video of the Presidential race (Hillary vs. Trump) had 47 million live views reaching 1.7 million viewers

Facebook Live Fails

Going viral has been a big part of internet culture. The more you have a “fail,” the more infamous you can be. Pakistan Minister’s cat filter fail is not the only incident of Facebook Live fails that has turned into a meme. In 2018, Ted Cruz broadcasted a Facebook Live with his video in portrait mode with his camera phone set up horizontally.  

Brands, even with their social media teams, are also not immune to the internet fails. Buzzfeed had a major fail back in 2016 when they had a Facebook Live interview with the then-president Barack Obama. After a lengthy introduction by host Chris Geidner, while waiting for the President, their live video feed stalled and failed to capture the interview which disappointing 35,000 Facebook viewers. The live stream of the event on Youtube, however, went well.

Consequently, there are now Youtube tutorials to help guide users on how to optimize Facebook Live and avoid these fails.

Aside from Facebook Live fails, the social media giant has been under fire since due to violent or disturbing live streams that are surfacing in the platform. Facebook has issued stricter restrictions to lessen the spreading of such videos.

CEO Mark Zuckerberg has called for support through his opinion piece in the Washington Post last May 30. He shared that Facebook, in collaboration with French officials, is creating an independent body that can help improve their content review systems. Through this initiative, Zuckerberg hopes that they can lessen, and in the future, eradicate harmful content.

Continue Reading

Editors' Pick

AMCA Breach: 20 Million Victims, 19 Class Actions

AMCA previously reported that there were only 200,000 victims in the recent breach, but their partner labs said there were more than 20 million. Click To Tweet

Published

on

AMCA previously reported that there were only 200,000 victims in the recent breach but the company, but their partner labs said there were more than 20 million.
AMCA previously reported that there were only 200,000 victims in the recent breach but the company, but their partner labs said there were more than 20 million. Photo: cbgrfx123 | Flickr | CC BY-SA 2.0

More than 20 million and not 200,000 have fallen victim to a massive data breach that has seen medical clients using the services of healthcare billing company, American Medical Collection Agency (AMCA), to pay for their laboratory tests in different blood testing labs across the U.S. were confirmed by the SEC filings of affected medical institutions amidst the earlier claims of AMCA that there were fewer victims.

The data breach was a result of a cyber attack that aims to phish for financial information from the website of the AMCA. The exposed data belongs to Americans who paid laboratory services at several clinical and blood testing labs and institutions and used the AMCA billing portal.

What happened in the AMCA breach?

Data that were stolen from the victims include their names, phone numbers, dates of birth, home addresses, social security numbers, credit card numbers, and other bank details. The said information was auctioned off by the hacker in several financial hacking forums.

According to DataBreaches.net, the organization who first reported about the incident, AMCA officials, following the notification of the breach confirmed that their system has been compromised and has remained undetected for more than eight months. AMCA corroborated that the breach took place between August 1, 2018, and March 30, 2019.

Notifications have been sent by several of AMCA’s corporate partners and clients to their customers following the disclosure of the security breach that has seen information from millions of Americans compromised.

Related: Hospitals Only Spend 5% Of Its Budget For Cybersecurity Amidst 82% Of Them Reporting To Have Been Attacked

The list of impacted testing laboratories includes Quest Diagnostics (11.9 million patients), LabCorp (7.7 million patients), BioReference Laboratories (Opko Health subsidiary, 422,600 patients), Carecentrix (500,000 patients), and Sunrise Laboratories (undisclosed number of patients).

However, neither the AMCA nor its five clients have yet to notify ALL impacted citizens by the breach making them vulnerable to a lot of cyber crimes and their financial data could be used by anyone who gets hold of the information against the persons of those who still don’t know that their financial information is floating around the internet.

The companies involved in the breach are facing several lawsuits

Appropriately, the lawsuit came into the direction of AMCA, Quest, and LabCorp regarding the incident. More than 11 class-suite actions have been filed against the three companies for their inability to protect consumer data. The 11 lawsuits were recorded at The United States Judicial Panel on Multidistrict Litigation (JPML) on June 3. Since then, eight more lawsuits were filed against the companies in federal courts from New Jersey, New York, and California.

According to litigation experts, “If many cases are filed in federal court, any of the lawyers on any of those cases can file a motion with the JPML [..] to centralize the various federal cases that have been filed by sending all of them to a single judge for coordinated pre-trial proceedings.”

“Healthcare companies are especially susceptible to data breaches not only because they aggregate a tremendous amount of important and sensitive data, but also because they tend to be less focused on cybersecurity protection than other industries,” said John Yanchunis of Morgan and Morgan, one of the firms who filed lawsuits against Quest Diagnostics.

“These companies, like Quest Diagnostics, know they are at an increased risk and yet have not taken the proper steps to protect their patients’ data. We will fight for justice on behalf of those impacted by this breach,” added Yanchunis.

Lawmakers are demanding an explanation

The U.S. government, led by attorneys general from Connecticut and Illinois has also opened an investigation on the matter. Furthermore, lawmakers and other politicians have sent letters to the responding companies to ask for an explanation of why an eight-month data breach remained undetected and to demand accountability from them.

In Washington, US Sen. Mark Warner (D-VA) also sent a letter to Quest Laboratories demanding the company explain its vetting process for selecting AMCA as a billing vendor, and what requirements a third-party vendor has to pass. Democratic New Jersey Sens. Cory Booker and Bob Menendez also sent letters to AMCA, Quest, and LabCorp, seeking official answers on how a breach of this severity went undetected for eight months.

“The months-long leak leaves sensitive personal and financial information vulnerable in the hands of criminal enterprises. Moreover, such breaches force victims to contend with identity theft that may lead to irreparable harm to their credit reports and financial future,” said the letter sent by the NJ senators.

Continue Reading

Technology

A Malware Has Been Pre-Installed In Some Cheap Android Devices, Google Confirms

Google confirms that some cheap Android devices have been found to have pre-installed malware known as Triada, which exploitis device root options. Click To Tweet

Published

on

Google confirms that some cheap Android devices have been found to have pre-installed malware known as Triada.
Cheap Android devices found out to have pre-installed malware. Photo: Rob Bulmahn | Flickr | CC By 2.0

In every purchase, sometimes you can get your money’s worth, sometimes you get something more. In a recent confirmation from Google, purchase of cheap and low-end Android devices also comes with pre-installed malware and has run undetected for a couple of years.

In a press release, Google, for the first time in history, has discussed in detail the malware that is called Triada, which the tech company has confirmed have been pre-installed in several low-end Android devices including Cherry Mobile, Leagoo, and Doogee. The malware, which was first discovered and published by Kaspersky Lab back in 2016 have been pre-installed in the affected devices, meaning, the malware already existed in the device even before someone buys it.

It was believed previously that the malware was added and installed to the affected devices at some point in the supply chain process. Now, Google has revealed that cybercriminals indeed managed to compromise Android smartphones and installed a backdoor while the supply chain process of the phones was underway.

Back in 2016, Triada was simply a rooting trojan that tried to exploit the device, and after getting elevated privileges, it performed a host of different actions. To hide these actions from analysts, Triada used a combination of dynamic code loading and additional app installs. According to the press release from Google, “Triada’s first action was to install a type of superuser (su) binary file. This (su) binary allowed other apps on the device to use root permissions.”

According to Google, Triad’s purpose is to install spam apps on a device by gaining root access. However, as Google’s security feature, Google Play Protect, improves in detecting malware, Triada was able to evolve to adapt to the new challenges posed by updated security firewalls by Google. Triada is known for downloading additional Trojan components on an infected device which then steals sensitive data from banking apps, intercepts chats from messengers and social media platforms, and there are also cyber-espionage modules on the device.

“The binary accepted two passwords, od2gf04pd9 and ac32dorbdq. This is illustrated in the IDA screenshot below. Depending on which one was provided, the binary either 1) ran the command given as an argument as root or 2) concatenated all of the arguments, ran that concatenation preceded by sh, then ran them as root. Either way, the app had to know the correct password to run the command as root,” Google said.

“This Triada rooting trojan was mainly used to install apps and display ads. This trojan targeted older devices because the rooting exploits didn’t work on newer ones. Therefore, the trojan implemented a weight-watching feature to decide if old apps needed to be deleted to make space for new installs.”

Affected devices

According to recent reports, the malware has affected over 40 devices. These devices include:

  • Leagoo M5
  • Leagoo M5 Plus
  • Leagoo M5 Edge
  • Leagoo M8
  • Leagoo M8 Pro
  • Leagoo Z5C
  • Leagoo T1 Plus
  • Leagoo Z3C
  • Leagoo Z1C
  • Leagoo M9
  • ARK Benefit M8
  • Zopo Speed 7 Plus
  • UHANS A101
  • Doogee X5 Max
  • Doogee X5 Max Pro
  • Doogee Shoot 1
  • Doogee Shoot 2
  • Tecno W2
  • Homtom HT16
  • Umi London
  • Kiano Elegance 5.1
  • iLife Fivo Lite
  • Mito A39
  • Vertex Impress InTouch 4G
  • Vertex Impress Genius
  • myPhone Hammer Energy
  • Advan S5E NXT
  • Advan S4Z
  • Advan i5E
  • STF AERIAL PLUS
  • STF JOY PRO
  • Tesla SP6.2
  • Cubot Rainbow
  • EXTREME 7
  • Haier T51
  • Cherry Mobile Flare S5
  • Cherry Mobile Flare J2S
  • Cherry Mobile Flare P1
  • NOA H6
  • Pelitt T1 PLUS
  • Prestigio Grace M5 LTE
  • BQ-5510 Strike Power Max 4G (Russia)

Reportedly, Leagoo and Cubot have already removed the malware from their affected devices since March 2018. Cherry Mobile also confirmed that they removed the malware from the affected devices in 2018.

Google also said that they worked with OEMs to remove the malware from devices and rolled out the fix through OTA updates.

“By working with the OEMs and supplying them with instructions for removing the threat from devices, we reduced the spread of preinstalled Triada variants and removed infections from the devices through the OTA updates,” said Lukasz Siewierski, Android Security & Privacy Team.

“The Triada case is a good example of how Android malware authors are becoming more adept. This case also shows that it’s harder to infect Android devices, especially if the malware author requires privilege elevation.”

Continue Reading

Trending